TL;DR

This project leverages LLM-based multi-agent systems to identify how networks of international cooperation enable authoritarian practices that transcend state borders, e.g.:

Problems with existing knowledge/methods

Substantive limitations

• Research on authoritarianism tends to focus on state actors and regime type
• Overlooks the role of other actors, e.g.: PR firms, technology companies, IOs
• ➡️ Authoritarian practices require networks of international cooperation to succeed, but we don’t know how these networks of international cooperation work

Methodological limitations

• Related work relies on manual data collection and structured data
• Existing approaches can’t capture interactions between different actors for empirical analysis

RQs

• RQ1: What types of actors are involved in enabling transnational authoritarian practices?
• RQ2: Through what mechanisms do different types of actors enable transnational authoritarian practices?
• RQ3: What do these patterns of cooperation reveal about how authoritarian practices are executed internationally?

Contributions

• Shifts the broader conversation on authoritarianism away from states/systems to practices, while providing a roadmap for analyzing the global scale/scope of said practices
  
• Theory: How different actor types engage in networks of international cooperation that enable authoritarian practices
• Data: Actor-level data on international cooperation in authoritarian practices
• Method: LLM-MAS as a way to analyze interactions between different types of actors involved in human rights practices/conflict processes/etc.

Appendix slide

Include visibility=“uncounted” in slide attributes if you want to make the progress bar end at the end of your main slides